Left ventricular flow propagation velocity measurement: Is it cast in stone?

Abstract

This study aims to investigate the measurement of left ventricular flow propagation velocity, V _p, using phase contrast magnetic resonance imaging and to assess the discrepancies resulting from inflow jet direction and individual left ventricular size. Three V _p measuring techniques, namely non-adaptive (NA), adaptive positions (AP) and adaptive vectors (AV) method, were suggested and compared. We performed the comparison on nine healthy volunteers and nine post-infarct patients at four measurement positions, respectively, at one-third, one-half, two-thirds and the conventional 4 cm distances from the mitral valve leaflet into the left ventricle. We found that the V _p measurement was affected by both the inflow jet direction and measurement positions. Both NA and AP methods overestimated V _p, especially in dilated left ventricles, while the AV method showed the strongest correlation with the isovolumic relaxation myocardial strain rate (r = 0.53, p < 0.05). Using the AV method, notable difference in mean V _p was also observed between healthy volunteers and post-infarct patients at positions of: one-half (81 ± 31 vs. 58 ± 25 cm/s), two-thirds (89 ± 32 vs. 45 ± 15 cm/s) and 4 cm (98 ± 23 vs. 47 ± 13 cm/s) distances. The use of AV method and measurement position at one-half distance was found to be the most suitable method for assessing diastolic dysfunction given varying left ventricular sizes and inflow jet directions.

Appendix: Determination of the angled velocity components

Referring to Fig. 7a, a reference line was extended from the mid-point of the MLT-inflow position to the apex point. Referring to the centre of the reference line as the pivot, the reference line was tilted at 1° intervals in the range of −45° to 45°.

Fig. 7

Procedure to obtain angled velocity components: a reference line and one of its tilted directions; b parallel velocity components, V _τ derived from velocity vectors and V at one-third measurement position

At each angle, an in-plane unit vector $\varvec{ }{\hat{\mathbf{\tau }}}$ parallel to the direction of the tilted reference line was acquired as in Eq. (3):

$$\hat{\varvec{\tau }} = \frac{{\overrightarrow {{{\mathbf{MA}}}} }}{{\left| {\overrightarrow {{{\mathbf{MA}}}} } \right|}} = \frac{{(x_{A} - x_{M} ), (y_{A} - y_{M} )}}{{\sqrt {\left( {x_{A} - x_{M} } \right)^{2} + \left( {y_{A} - y_{M} } \right)^{2} } }} = \left( {\tau_{1} ,\tau_{2} } \right)$$

(3)

At each pixel along the MLT-inflow or at each measurement position, the velocity component V_τ parallel to the tilted reference line was evaluated according to Eq. (4):

$$V_{\tau } = {\mathbf{V}} \cdot \hat{\varvec{\tau }} = V_{1} \tau_{1} + V_{2} \tau_{2}$$

(4)

where V = (V ₁, V ₂) is the in-plane velocity vector at the given pixel. Figure 7b further illustrates the physical interpretation of ${\mathbf{V}}_{\varvec{\tau}} \left( { = V_{t} \hat{\varvec{\tau }}} \right)$) at the one-third measurement position. Due to the choice of the direction of the unit vector, forward flow (flow towards the apex) yields positive values for V_τ. Only these were used to obtain the mean forward flow.